immunoPETE: A DNA-based integrated B-cell and T-cell receptor profiling platform

This paper introduces and comprehensively benchmarks immunoPETE, a robust DNA-based assay that enables simultaneous, accurate, and quantitative profiling of B-cell and T-cell receptor repertoires from genomic DNA, demonstrating its utility in predicting clinical outcomes in bladder cancer and its broad applicability across immunology research.

The Gist

Imagine your immune system as a massive, highly organized library. Inside this library, there are billions of unique books (cells) that teach your body how to fight off specific invaders like viruses, bacteria, or even cancer. Each book has a unique "title" and "author" (its genetic code) that determines exactly what enemy it can recognize.

For a long time, scientists could only read a few pages of these books, or they had to wait until the books were actively being "read" (transcribed into RNA) to see what was happening. This was like trying to understand a library by only looking at the books currently being checked out, missing the vast archive of books sitting on the shelves.

Enter "immunoPETE": The Ultimate Library Catalog.

This paper introduces a new tool called immunoPETE. Think of it as a super-powered scanner that can instantly catalog every single book in your immune library, directly from the DNA (the permanent record) rather than waiting for the books to be read.

Here is how it works, broken down into simple concepts:

1. The "One-Stop-Shop" Scanner

Traditionally, if you wanted to know about your "B-cell" books (which make antibodies) and your "T-cell" books (which hunt infected cells), you had to run two separate, expensive, and time-consuming scans.

• The Analogy: It's like having to go to the fiction section to catalog novels and then walking all the way to the non-fiction section to catalog biographies, using two different machines.
• The immunoPETE Solution: This tool is a "universal scanner." It can scan both the B-cell and T-cell sections simultaneously in a single reaction. It even catches the "Gamma-Delta" T-cells, which are like the special forces of the immune system that often get ignored in standard scans.

2. The "Name Tag" System (UMIs)

One of the biggest problems in scanning DNA is that the process of copying the DNA (amplification) can sometimes create errors or make one copy look like a thousand.

• The Analogy: Imagine trying to count how many people are at a concert by asking them to shout their names. If one person shouts twice, you might think there are two people. Or if the microphone is bad, you might miss someone.
• The immunoPETE Solution: Before scanning, the tool attaches a unique "name tag" (called a UMI) to every single original DNA molecule. It's like giving every person at the concert a unique wristband. Even if the system copies the DNA a million times, it knows that all those copies came from the same original person. This allows the scientists to count the exact number of unique immune cells with incredible precision, correcting for any "shouting" errors.

3. Testing the Scanner

The authors didn't just build the scanner; they put it through the wringer to prove it works.

• The "Fake Crowd" Test: They created artificial mixtures of known cell types (like mixing 50% red marbles and 50% blue marbles) to see if the scanner could accurately report the ratio. It did, with near-perfect accuracy.
• The "Needle in a Haystack" Test: They tried to find a single specific cell type hidden among millions of others (like finding one specific needle in a giant haystack). The scanner could find this "needle" even when it made up only 0.01% of the total mix. This is crucial for catching early signs of cancer or disease.
• The "Repeatability" Test: They scanned the same sample eight times. The results were almost identical every time, proving the tool is reliable and doesn't give random answers.

4. Real-World Application: The Bladder Cancer Case Study

To show how this helps real patients, the team looked at people with non-muscle-invasive bladder cancer.

• The Problem: Usually, to study the immune system inside a tumor, doctors have to do a painful biopsy (cutting out a piece of the tumor).
• The immunoPETE Discovery: They tested urine samples. Because the bladder is a hollow organ, immune cells shed from the tumor into the urine.
• The Result: The "library catalog" from the urine looked almost exactly like the catalog from the actual tumor tissue! This means doctors might soon be able to monitor a patient's immune response to cancer just by analyzing a urine sample, avoiding painful surgeries.
• The Bonus: When they added these immune "library stats" to the standard medical data (like age and tumor size), they could predict which patients were likely to get worse much better than using medical data alone.

Why Does This Matter?

Think of immunoPETE as a high-resolution map of your immune system's history and current status.

• For Cancer: It helps find the "bad clones" (cancer cells) hiding in the crowd and predicts if a patient will respond to treatment.
• For Infections: It can track how your body remembers a virus (like COVID or the flu) over time.
• For Autoimmune Diseases: It can spot when your immune system is writing "books" that attack your own body.

In short, immunoPETE turns the chaotic, invisible world of your immune cells into a clear, readable, and actionable map, helping doctors make smarter decisions about how to treat diseases like cancer, infections, and autoimmune disorders. It's a move from "guessing" what's happening inside you to "knowing" exactly what's happening.

Technical Summary

1. Problem Statement

The adaptive immune system relies on the vast diversity of B-cell receptors (BCRs) and T-cell receptors (TCRs) generated through V(D)J recombination. Profiling these immune repertoires is crucial for understanding immune responses in cancer, infection, and autoimmunity. However, existing technologies face several limitations:

• RNA-based bias: Traditional RNA-seq assays rely on mRNA transcription levels, which can vary significantly between cell states (e.g., resting vs. activated plasma cells), leading to inaccurate quantification of clonal abundance.
• Fragmented profiling: Most assays target only a single chain (e.g., TRB or IGH) or require separate reactions for B and T cells, preventing simultaneous, internally normalized analysis of B-T interactions.
• Lack of single-cell resolution: Without Unique Molecular Identifiers (UMIs), it is difficult to distinguish true biological clones from PCR/sequencing errors or to achieve accurate cell-level quantification.
• Sample limitations: Many methods struggle with low-input samples (e.g., urine, FFPE) or fail to capture rare clones necessary for Minimal Residual Disease (MRD) monitoring.

2. Methodology: The immunoPETE Assay

immunoPETE is a genomic DNA (gDNA)-based assay designed to simultaneously profile human TRB, TRD, and IGH chains in a single reaction.

• Workflow:
  1. Input: Uses genomic DNA extracted from various sources (PBMCs, whole blood, FFPE, urine, CSF).
  2. Two-Stage Primer Extension:
     • Stage 1: A pool of V-gene oligos containing an 11-base Unique Molecular Identifier (UMI) and a universal sequence performs primer extension on gDNA. Exonuclease treatment removes unextended DNA and excess oligos.
     • Stage 2: A master mix containing J-gene oligos and sequencing adapters (i7/i5) performs a second extension.
  3. Library Prep: Libraries are purified, quantified, and sequenced on Illumina platforms (typically 2x150 bp paired-end).
  4. Bioinformatics (Daedalus Pipeline):
     • Reads are merged, trimmed, and aligned to IMGT references.
     • UMI-based clustering: Reads are grouped into "UMI families" based on CDR3 and UMI identity.
     • Consensus generation: Errors are corrected by generating consensus sequences within families.
     • Filtering: Functional rearrangements are identified based on non-chimeric V/J annotations, CDR3 length (>3 AA), and absence of stop codons/frameshifts.
  5. Output: Cell-level resolution data including CDR3 sequences, V/J gene usage, clone counts, and repertoire diversity metrics.

3. Key Contributions

• Integrated Multi-Chain Profiling: First platform to simultaneously capture TRB, TRD (including $\gamma\delta$ T cells), and IGH in a single reaction, enabling internal normalization of B/T cell ratios.
• DNA-Based Quantitation: By targeting gDNA (1 template = 1 cell), it avoids transcriptional biases inherent in RNA-based methods, providing a stable snapshot of the total lymphocyte population (active + bystander).
• UMI-Driven Accuracy: Integration of UMIs allows for error correction and precise reconstruction of full-length amplicon consensus sequences, enabling true cell-level resolution.
• Comprehensive Benchmarking: The study provides a rigorous analytical validation including Limit of Blank (LoB), Limit of Detection (LoD), Limit of Quantitation (LoQ), linearity, and reproducibility across diverse sample types.

4. Key Results

• Accuracy & Reproducibility:
  • Sequence Fidelity: Using the monoclonal HuT-78 cell line, the assay achieved >98% purity in clonal sequence retrieval at inputs $\ge$ 2 ng.
  • Cell Type Ratios: In contrived Pan-T/Pan-B mixtures, the assay recovered B/T ratios with a Mean Absolute Error (MAE) $\le$ 3% and a Coefficient of Variation (CV) $\le$ 0.15.
  • Reproducibility: High-complexity PBMC samples showed consistent cell recovery and clonal composition across 8 technical replicates (CV $\le$ 0.15 for chain fractions). Top clone abundances were highly reproducible even for clones representing <1% of the repertoire.
• Quantitative Limits:
  • Linearity: Strong linear correlation ($R^2 > 0.98$) between DNA input mass (500–2500 ng) and cell recovery. Linearity extended down to 2 ng input on a log scale.
  • Sensitivity: The Limit of Detection (LoD) was established at 51.6 cells, and the Limit of Quantitation (LoQ) at 122.3 cells.
  • Rare Clone Detection: The assay detected spiked HuT-78 clones in PBMC backgrounds with 100% sensitivity at $10^{-3}$ and $10^{-4}$ fractions, and 75% sensitivity at $10^{-5}$ (0.02 ng input).
• Gene Usage & Pairing:
  • Demonstrated that V and J gene pairing is non-random (Chi-squared test, $p \le 10^{-32}$), with the strongest non-randomness observed in the TRD chain.
• Clinical Case Study (NMIBC):
  • Urine Profiling: Successfully generated immune repertoires from urine pellets, showing that urine immune profiles more closely resemble tumor tissue than peripheral blood (higher Jaccard similarity).
  • Prognostic Value: In a cohort of 24 non-muscle invasive bladder cancer (NMIBC) patients, adding immune biomarkers (derived from blood repertoires) to clinical factors improved the prediction of progression-free survival (Concordance score increased from 0.625 to 0.689).
  • Biomarker Discovery: Identified a trend where lower TRDV1 gene usage correlated with higher risk of disease recurrence/progression.

5. Significance

• Precision Medicine: immunoPETE offers a robust, quantitative tool for monitoring clonal evolution in cancer (e.g., MRD), tracking vaccine responses, and studying autoimmune diseases.
• Non-Invasive Diagnostics: The ability to profile immune repertoires from urine and FFPE tissues opens new avenues for non-invasive liquid biopsies and retrospective archival studies.
• Comprehensive Immune View: By capturing $\gamma\delta$ T cells and B-cell dynamics alongside $\alpha\beta$ T cells in a single workflow, it provides a more holistic view of the immune microenvironment, particularly in the tumor microenvironment where these "marginal" populations play critical roles.
• Scalability: The assay has been used in large-scale projects like the Human Blood Immunome Encyclopedia (HuBIE), demonstrating its potential for population-level immune monitoring and machine-learning-based diagnostic development.

In summary, immunoPETE represents a significant advancement in immune repertoire sequencing by combining DNA-based stability, UMI-driven accuracy, and multi-chain integration, making it a powerful tool for both research and future clinical applications in oncology and immunology.